Sei sulla pagina 1di 27

International Journal of Bank Marketing

Digital banking, customer experience and bank financial performance: UK


customers’ perceptions
Cajetan I. Mbama, Patrick O. Ezepue,
Article information:
To cite this document:
Cajetan I. Mbama, Patrick O. Ezepue, (2018) "Digital banking, customer experience and bank
financial performance: UK customers’ perceptions", International Journal of Bank Marketing, Vol. 36
Issue: 2, pp.230-255, https://doi.org/10.1108/IJBM-11-2016-0181
Permanent link to this document:
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

https://doi.org/10.1108/IJBM-11-2016-0181
Downloaded on: 09 October 2018, At: 02:01 (PT)
References: this document contains references to 81 other documents.
To copy this document: permissions@emeraldinsight.com
The fulltext of this document has been downloaded 1972 times since 2018*
Users who downloaded this article also downloaded:
(2017),"Building customer loyalty in digital banking: A study of bank staff’s perspectives on
the challenges of digital CRM and loyalty", International Journal of Bank Marketing, Vol. 35 Iss
6 pp. 858-877 <a href="https://doi.org/10.1108/IJBM-08-2016-0112">https://doi.org/10.1108/
IJBM-08-2016-0112</a>
(2017),"Mobile banking service quality and customer relationships", International Journal of Bank
Marketing, Vol. 35 Iss 7 pp. 1068-1089 <a href="https://doi.org/10.1108/IJBM-10-2015-0150">https://
doi.org/10.1108/IJBM-10-2015-0150</a>

Access to this document was granted through an Emerald subscription provided by emerald-
srm:380143 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald
for Authors service information about how to choose which publication to write for and submission
guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as
well as providing an extensive range of online products and additional customer resources and
services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the
Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for
digital archive preservation.

*Related content and download information correct at time of download.


The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0265-2323.htm

IJBM
36,2 Digital banking, customer
experience and bank financial
performance
230 UK customers’ perceptions
Received 30 November 2016
Revised 1 March 2017
Cajetan I. Mbama and Patrick O. Ezepue
Accepted 28 March 2017 Sheffield Hallam University, Sheffield, UK

Abstract
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

Purpose – The purpose of this paper is to examine customers’ perceptions of digital banking (DB), customer
experience, satisfaction, loyalty and financial performance (FP) in UK banks.
Design/methodology/approach – The research consists of a survey of UK bank customers’ perceptions of
the above themes; use of banks’ financial reports to obtain FP ratios; multivariate factor analysis; structural
equation modelling; and analysis of variance tests to explore research hypotheses on the relationships among
the study factors.
Findings – The main factors which determine customer experience in DB are service quality, functional
quality, perceived value (PV ), employee-customer engagement, perceived usability and perceived risk. There
is a significant relationship among customer experience, satisfaction and loyalty, which is related to FP.
Research limitations/implications – This study concentrates on UK bank customers which limits its
generalisability to other banks globally. However, the fact that banks typically adopt common standards in
bank financial management implies that the findings are potentially robust for global bank management.
Replicating the study in banks in other countries will further enhance this robustness.
Practical implications – Some significant effects of customer characteristics on the study factors were
observed, which have useful implications for DB, bank marketing services and bank FP.
Originality/value – Unlike previous studies, this study uses both Net Promoter Score and financial ratios as
dependent variables, to provide a combined study of the relationships among 14 study factors, with
implications for bank marketing and FP.
Keywords Service quality, Financial performance, Customer satisfaction, Customer loyalty,
Customer experience, Digital banking
Paper type Research paper

1. Introduction
The development of technology in the banking sector has significant implications for banks’
marketing efforts (Dootson et al., 2016), especially in digital banking (DB) as it affects
customer interfaces. DB via telephone, internet and mobile has become a major way of
delivering multi-channel services to customers, which is challenging traditional banking
models (Cortiñas et al., 2010). As customers’ expectations increase, capturing and retaining
them and improving profitability becomes important, especially after the financial crisis of
2008 (Monferrer-Tirado et al., 2016).
Moreover, increasing DB uptake has made some UK banks reduce branch numbers
(BBC, 2016; French et al., 2013). This shift towards DB means that banks’ marketing and
financial management models are changing, making it crucial for banks to understand the
impact of DB on customer experience and financial performance (FP) phenomena.
This paper, therefore, fills the need for a composite understanding of UK customers’
perceptions of the relationships among DB, customer experience and bank FP, and their
implications for bank marketing. This perspective has not been explored in the literature.
International Journal of Bank
Marketing Some previous studies focussed on marketing and service quality, the relationship
Vol. 36 No. 2, 2018
pp. 230-255
among customer satisfaction and loyalty ( Jun and Palacios, 2016; Amin, 2016), FP of banks
© Emerald Publishing Limited
0265-2323
(Keisidou et al., 2013; Liang et al., 2009) and mainly customer experience (Klaus and Maklan,
DOI 10.1108/IJBM-11-2016-0181 2013; Garg et al., 2014), giving limited attention to the effect of customer experience on FP.
Recent research has investigated internet and mobile banking service quality and customer UK customers’
satisfaction (Amin, 2016; Saleem et al., 2016), but DB is not yet treated holistically in bank perceptions
marketing as pursued in this paper. Additionally, Piyathasanan et al. (2015), in their study of
the effects of internet experience on customer value perception, argue that few guidelines
are available on how to improve consumers’ digital experience.
Evidently, banks are service providers whose financial success depends on customers’
perceived service quality and experience (Andaleeb et al., 2016). The service marketing 231
challenge for banks is to overcome customers’ reluctance to use DB due to bad
experience. Developing service marketing theory for DB requires an understanding of
customers’ preferences, and drivers of customer satisfaction and loyalty (Grönroos, 1984),
as in this study. Although DB is ubiquitous among all mainstream banks, there is still a
need to understand the impact of DB on bank FP, especially for different customer
segments (Keisidou et al., 2013; Patsiotis et al., 2012; Garg et al., 2014). This will enable
different banks to fine-tune their bank marketing strategies in line with their overarching
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

business model.
This paper presents an integrated understanding of customers’ perceptions of the links
among DB, customer experience, satisfaction, loyalty, two measures of FP, namely
traditional financial ratios (FP1) and Net Promoter Scores (NPS) criteria (FP2), and their
implications for bank marketing. The specific research objectives are:
(1) to perform exploratory data analyses (e.g. descriptive analysis and paired
correlations) of the research data on 14 key factors (see Figure 1) which relate to DB,
customer experience, satisfaction, loyalty and the FP1-FP2 measures;
(2) to use exploratory factor analyses to test the strengths of the relationships among
the factors, and their associated loadings on questionnaire items, with a primary
focus on how these results underpin FP1 and FP2;
(3) to apply confirmatory structural equation modelling (SEM) of these relationships to
explore some subsidiary and primary hypotheses on the extent and direction of

Perceived value
0,
H1 em1
1
Convenience

H2 Customer satisfaction
0,
Functional quality em4 0,
H3 H13 em3
1 H10 1
Digital BK service quality H15
H4
H12 Financial performance
Customer experience
H5 0,
Brand/trust
em2
H11 1 H14
H6
Employee customer Egt

H7 Customer loyalty

Perceived risk
H8

Perceived usability

H9 Figure 1.
Digital banking innovation
Conceptual model
IJBM relationships among the explanatory variables, and triangulate the relationships in
36,2 the factor analyses; and
(4) to use analysis of variance (ANOVA) tests to determine which customer
characteristics are significantly associated with the study factors, which is crucial
for using the research findings in bank marketing.
The rest of the paper consists of a critical literature review (Section 2); a conceptual model
232 linking the study variables (Section 3); methodology (Section 4), measurement development
including content validity (Section 5); empirical analysis and results by specific research
objectives (Section 6); discussion of the theoretical and managerial implications of the results
(Section 7); and conclusion, limitations and suggestions for future work (Section 8).

2. Literature review
2.1 DB
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

The use of technology such as DB in service innovation to meet client needs is best
understood through its relationship to the service users and how they perceive the service
(Baba, 2012). A theory in marketing studies is a logically self-consistent model that explains
how related phenomena behave (Lee and Greenley, 2008). Marketing theory and models
explore how some intrinsic and extrinsic factors shape customers’ service perceptions
and firms’ profitability (Grönroos, 1982); for example, Service Profit Chain (SPC) model
(Heskett et al., 2008) and NPS (Reichheld, 2003) indicate that improving customer service
attributes can improve profitability. Davis (1989) postulates that perceived ease of use and
usefulness factors influence customers’ behaviour in using new technology. In Jordanian
banks, perceived usefulness, trust and self-efficacy are the predictors of customers’ use of
telebanking (Alalwan et al., 2016), but generally in DB experience contexts, different factors
may be applicable, which need to be explored.
Telephone, internet and mobile have become major DB service channels, making them
important for banks’ survival, through the advantages of convenience, and anytime,
anywhere service access (Sundarraj and Wu, 2005; Daniel, 1999; Mols, 2001). Some
researchers argue that e-banking services facilitate good customer services, which retain
customers (Martins et al., 2014). These three digital devices offer different interfaces and
choices to customers, with telephone banking being the earliest. There have been
fragmented studies in DB, with authors studying individual channels or focusing attention
on only few variables among those explored in this paper. Amin (2016) and Raza et al. (2015)
study internet banking service quality and its relationship to customer loyalty, while Jun
and Palacios (2016) study mobile banking service quality. Although this type of study offers
advantages in certain contexts, a comprehensive study is required to understand customers’
general view of DB experience and FP.
DB enables banks to develop services for customers, cut costs associated with sending
statements by post and face-to-face transactions with customers in branches. Nowadays,
customers expect to have similar levels of interactions in DB and social media (Dootson et al.,
2016). As customers increasingly accept DB, more than 600 UK bank branches have closed,
with rural areas worst affected (BBC, 2016; French et al., 2013). The impact of DB becoming
dominant on customer experience, operational efficiency and FP and marketing are still not
clear, hence this paper.
Although DB is technology related, it is service-oriented, making service marketing
theories important in its conceptualisation (Van Looy et al., 1998). Previous studies are
focussed on factors influencing users’ intentions to adopt internet and mobile banking
respectively (Martins et al., 2014 and Zhou, 2012). These studies relate more to DB
acceptance than experience and may not fulfil the current banking marketing needs.
Alternative studies note that internet banking derives from unique service and functional
qualities (Kaura et al., 2015; Monferrer-Tirado et al., 2016). Similarly, Lee and Chung (2009) UK customers’
indicate that good user interface quality affects trust in and satisfaction with mobile perceptions
banking. These perspectives are investigated further through customers’ opinions.
Consequently, Hoehle et al. (2012) note that whilst the utilisation of DB channels has
grown substantially, prior study has not identified all the customer-related issues, and may
be limited due to fragmented findings and methods of study. Hence, this paper presents a
more comprehensive study of the landscape of UK DB linked to customer experience, 233
FP and service marketing. This approach supports more robust theory development than
was possible in previous studies.

2.2 Customer experience and FP


Different theories and models for measuring customer satisfaction and organisational
performance have emerged (e.g. NPS, Reichheld, 2003; SERVQUAL, Parasuraman et al.,
1988; and SPC, Heskett et al., 2008). The SPC model establishes the relationships between
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

service quality, employee job satisfaction (employee retention and productivity), customer
satisfaction and loyalty and organisational performance (revenue growth and profitability).
Kanyurhi and Akonkwa (2016) used the SPC model in Congo banks and found a positive
relationship between internal marketing and employee satisfaction, and a positive
relationship between internal marketing and perceived organisational performance, but not
between employee satisfaction and perceived organisational performance.
The NPS gauges the level of customer satisfaction and loyalty to a firm, using a single
question, while SERVQUAL measures service qualities using reliability, tangibles
assurance, responsiveness and empathy, without demonstrating their direct relationship
to profitability. In DB, new models are required, as customer priorities in contact services
may not be applicable, for example courtesy, friendliness and personal care. Hence, new
service quality measures that moderate customer satisfaction in DB have emerged
( Jun and Palacios, 2016; Amin, 2016; Dootson et al., 2016). Since this study utilises
some service quality measures (experience, satisfaction and loyalty), it will contribute
new knowledge on any significant relationships they may have with digital bank
marketing and FP.
Customer experience encompasses a set of interactions (e.g. rational, emotional, sensorial,
physical and spiritual) between a customer, product and company, the value created
through that set of interactions (Meyer and Schwager, 2007; Verhoef et al., 2009; Klaus and
Maklan, 2013), and customers’ purchasing behaviour (Klaus and Maklan, 2013). Customers
compare their service expectations and their experiences interacting with firms’ offerings
during different service contacts. By focussing on a few factors at a time, previous studies
do not provide a strong enough evidence base for constructing more robust theories of
the links among customer experience attributes, different FP measures (FP1 and FP2), and
bank marketing.
With regards to customer experience measures, Meyer and Schwager (2007) advocate the
use of NPS, which captures the net result of good experience minus bad experience of what
customers know about a firm. They conclude that customer satisfaction occurs when the
gap between customers’ expectations and experiences has been closed. Thus, banks should
constantly seek the opinion of customers about their DB to improve their experience.
The above-mentioned link between customer experience and purchasing behaviour
suggests that customer experience is mediated by marketing to improve customer
satisfaction and loyalty, and their impacts on an organisation’s FP. As argued above, these
aspects of DB have not received detailed attention in the literature on bank marketing and
financial management, hence this paper.
Importantly, Maklan and Klaus (2011) recommend that researchers should explore which
dimensions of customer experience are important for organisational performance. This will
IJBM enable bank marketing to maximise FP, through customer experience, loyalty and
36,2 satisfaction and share-of-wallet. Their study was conducted in contact services and may not
fulfil the needs of DB. However, it reinforces the need to explore significant relationships
among customer characteristics and the various explanatory and dependent variables used
in this study, since these clarify customer experiences that will support successful digital
bank marketing (see Objective 4 above).
234 Most research in customer experience explores consumer perceptions (Holbrook, 2000)
and customer experience management (CEM) (Schmitt, 2003). Berry et al. (2002) suggest that
the first step in CEM should be defining all the clues that a firm communicates to customers,
to determine whether the company is meeting them. The findings in this paper could
provide clues on interface design, functionality, usability and quality of service, from
customers’ experience of DB, the areas that have previously received limited attention.
Linking customer experience to FP. Improved customer experience can offer values to
both firms and customers, such as enhanced customer satisfaction and loyalty to
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

organisations’ offerings, positive word-of-mouth referrals (e.g. captured by NPS), improved


retention, reduced complaints and fines, all of which can improve profitability (Reichheld
et al., 2000). In sum, using NPS categories to link customer experience (satisfaction, loyalty
and behavioural intentions, such as how strongly customers will recommend products to
others) to profit, has become popular in measuring organisational performance, but has not
been widely replicated in DB. Dootson et al. (2016) note that perceived usefulness, economic
value and social value predict overall perceived value (PV ), which in turn predicts a
customer’s intention to use social media to interact with a bank. However, a major limitation
of previous research in DB is showing how this type of value improves FP, instead of a
predominant focus on service quality, satisfaction and loyalty ( Jun and Palacios, 2016;
Amin, 2016). This paper fills this gap in knowledge through DB in the UK.
Recent studies on satisfaction, loyalty and FP concentrate attention on traditional
banking. Keisidou et al. (2013) found no relationship between customer satisfaction, loyalty
and FP in Greek banks, but Chi and Gursoy (2009) found a relationship between customer
satisfaction and FPFP in the US hospitality sector. Liang et al. (2009) found that the impact
of product attributes on customer satisfaction, trust/commitment and customer loyalty, and
FP in Taiwanese banks, again not in DB contexts. Given the conflicting evidence of the links
between customer experience variables and FP in these various contexts, this paper
presents much-needed evidence of these effects in UK DB.
Measuring customer experience and FP. Brakus et al. (2009) developed a four-dimensional
brand experience scale using measures such as sensory, affective, intellectual and
behavioural, highlighting the positive relationship between brand experience, brand
personality, satisfaction and loyalty. However, Novak et al. (2000) propose online customer
experience instruments with constructs such as web usage, arousal, challenge, control,
exploratory behaviour, flow, focussed attention, interactivity, involvement, playfulness,
positive effect, skill, tele-presence and time distortion. Although these instructions are useful,
some do not relate to DB (e.g. playfulness). Garg et al. (2014) investigate customer experience
in their model and found convenience to be the most important factor, followed by customer
interaction, employees, speed, servicescape, core service, online functional elements, presence
of other customers, value addition, service process and online aesthetics, while the marketing-
mix, customisation and online hedonic factors are moderately significant. These findings
relate to a prior study (Al-Eisa and Alhemoud, 2009), that focussed on measuring customer
satisfaction in Kuwait banks. Garg et al. (2014) consider both online and offline activities (e.g.
servicescape), which relate to the physical environment of services, while DB experience
should involve direct customer interaction through the online interface. This study considers
factors which are related to online activities only.
Researchers tend to study customer experience with customer satisfaction and loyalty; UK customers’
Saleem et al. (2016) note that social influence, market orientation and service quality relate to perceptions
customer loyalty, with satisfaction as an antecedent, and advise that banks should invest
resources to enhance customer satisfaction and loyalty. Similarly, Akhter et al. (2011)
indicate that customer loyalty is positively related to customer satisfaction, product image,
trustworthiness and customer relationship. Both studies were carried out in Pakistan banks,
which is different from the UK context for which the factors affecting satisfaction and 235
loyalty can vary. Fathollahzadeh et al. (2011) study the online and offline effects of
satisfaction, co-operation, trust, commitment, service quality, complaint handling, image
and communication in the Iranian banking sector and find that all eight variables have a
significant relationship with customer satisfaction, which can lead to customer loyalty.
Nevertheless, customer satisfaction is an aggregate of past consumption experience, and
gives a limited prediction of future customer behaviour, while loyalty offers a better
predictor of repeat purchasing intention (Liang et al., 2009). Thus, both variables are studied
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

in this paper, since they offer different customer perspectives.


For studies that combine a number of factors, Klaus and Maklan’s (2013) measurements of
customer experience comprise six measures, namely convenience, retention, service recovery
(part of service quality), risk perception, satisfaction and loyalty intentions, while Liang et al.
(2009) use three factors which are antecedents of trust, loyalty and FP. These authors’
research instruments are not all related to DB. Keisidou et al. (2013) investigate the
relationship between three factors: customer satisfaction, loyalty and FP. They used financial
ratios to measure FP of banks (e.g. return on assets (ROA) or investment (ROI), net profit
margin (NPM) and return on equity (ROE)). Methodologically, some of the above measures are
not commensurate across the banks when they are not stated in percentages, which is a
limitation; therefore, this study uses percentage measures of ROE, net interest margin (NIM)
and cost-to-income ratio. Chang and Lin (2015) developed a customer experience framework in
the Taiwanese leisure industry, using experiential value which customers derive from
services. This study incorporates PV in the analysis of customer perceptions of UK DB.
In sum, previous studies have not sufficiently integrated customer experience and FP in DB.
Instead of the limited factors typically used in previous studies, this paper uses 14 key factors
to provide a more detailed understanding of their relationships and impact on bank marketing
and FP. As argued above, this more holistic perspective has not been attempted in previous
studies. The following conceptual model summarises the selected variables and notations used.

3. Conceptual model for the research


Figure 1 illustrates the research conceptual model. In the model, “em” symbolises the
dependent variable while the rest are independent variables, “BK” means banking, “Egt”
engagement and H is hypothesis. Many of these hypotheses are secondary exploratory
hypotheses, whilst the primary research hypotheses relate to the links among customer
experience (satisfaction and loyalty), bank FP and marketing.
In Figure 1, nine independent variables are linked to customer experience through nine
exploratory hypotheses, H1-H9; customer experience and the remaining three dependent
variables – customer satisfaction, loyalty and FP (FP1 and FP2 measures) linked through
six primary hypotheses, H10-H15. The model encompasses 14 factors gleaned from the
literature to be appropriate for DB and 15 testable hypotheses. Further rationale for the
hypotheses development is provided below.

3.1 Research hypotheses development


Perceived value. PV is defined as the trade-off between costs and benefits of performing a
behaviour (Dootson et al., 2016). It is an important determinant of behavioural intentions to use
IJBM e-commerce (Ko et al., 2009; Piyathasanan et al., 2015), and a key factor for banks (Keisidou
36,2 et al., 2013; Liang et al., 2009; Garg et al., 2014; Fathollahzadeh et al., 2011) and the leisure
industry (Chang and Lin, 2015). However, limited attention has been given to it in UK DB.
We therefore frame an exploratory hypothesis linking PV and customer experience, namely:
H1. There is a positive relationship between PV and customer experience.

236 Convenience. Convenience is rarely researched alongside customer experience; however, it


has a positive effect on customer satisfaction (Keisidou et al., 2013; Knutson et al., 2007; Kim
et al., 2011; Karatepe et al., 2005) and customer experience (Garg et al., 2014; Klaus and
Maklan, 2013) relating to offline and online activities. Jun and Palacios (2016) see
convenience as one of the key service qualities of mobile banking in the USA. Wu (2011)
investigated the location convenience effect on customer satisfaction, while Keisidou et al.
(2013) tested the operational and locational characteristics of convenience. In this study,
operational convenience of DB is tested:
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

H2. There is a positive relationship between convenience and customer experience.


Functional quality. This deals with the functionality aspect of online systems, its activities and
interactivity components, which affects customer experience (Garg et al., 2014). Functional
quality influences users’ uptake of mobile banking (Lee and Chung, 2009) and customer
satisfaction (Keisidou et al., 2013), and customer satisfaction and trust, and loyalty in Spanish
banks (Monferrer-Tirado et al., 2016). These effects need to be tested in UK banks:
H3. There is a positive relationship between functional quality and customer experience.
Service quality. Service quality is widely tested in banks, hotels and insurance companies
employing SERVQUAL (Parasuraman et al., 1988), which differs slightly in DB. In banking,
service quality is found to increase customer satisfaction (Keisidou et al., 2013; Kaura et al.,
2015) and profitability (Ladhari et al., 2011). It mediates overall satisfaction, which is an
antecedent of loyalty (Levy and Hino, 2016). Jun and Palacios (2016) study mobile banking
service quality in USA, while Amin (2016) and Raza et al. (2015) study internet banking
service quality and its relationship to customer satisfaction and loyalty in Saudi Arabia and
Pakistan, respectively. The authors found that service quality significantly impacts on
satisfaction and consequently leads to loyalty. However, there is limited research on service
quality as a concept in DB services marketing in UK, hence the following hypothesis:
H4. There is a positive relationship between digital bank service quality and customer
experience.
Brand trust. Brand credibility is the level at which the service proposition information is
considered to be believable (Keisidou et al., 2013). Brand, trustworthiness and image have
been researched in studies and found to affect customers’ bank choices (Liang et al., 2009;
Fathollahzadeh et al., 2011; Knutson et al., 2007; Akhter et al., 2011). Levy and Hino’s (2016)
study found that attachment to brand positively affects bank loyalty. This study explores
brand trust in UK banks via the hypothesis:
H5. There is a positive relationship between brand trust and customer experience.
Employee-customer engagement (ECE). Bank employees have interactions with customers
and are the most important link in service delivery and complaint handling processes
(Karatepe and Aga, 2016). They need to be friendly, competent, capable of sustaining
interpersonal distance (Garg et al., 2014; Verhoef et al., 2009), and in building trust and
influencing customer behaviour. Employees’ attributes influence firms’ business
performance in the non-DB environment (Grace and O’Cass, 2004; Karatepe et al., 2005),
and customer satisfaction and profit (Yee et al., 2010). Customer satisfaction is largely
influenced by service quality, which depends on employee job satisfaction in contact UK customers’
services (Chi and Gursoy, 2009; Kanyurhi and Akonkwa, 2016). However, the relevance of perceptions
ECE in DB experience is tested:
H6. There is a positive relationship between ECE and customer experience.
Perceived risk. Jun and Palacios (2016) found security as one of the key factors that affects
service quality of mobile banking, while perceived risk is found to affect DB uptake in other 237
countries (Martins et al., 2014; Akinci et al., 2003; Hanafizadeh et al., 2014). Banks are
constantly investing in security to minimise risks; the following hypothesis is explored:
H7. There is a negative relationship between perceived risk and customer experience.
Perceived usability. DB is about customers’ electronic interface with a bank. Usability is one of
the key elements that determines mobile banking uptake (Gu et al., 2009), and e-commerce
business customer experience (Klaus, 2013). A study of Jordanian banks found that perceived
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

usefulness, trust and self-efficacy are predicting factors to use of telebanking (Alalwan et al.,
2016). This is tested in the context of DB experience, thus it is hypothesized that:
H8. There is a positive relationship between “perceived usability” and customer experience.
DB innovation. Banks benefit from interactive service innovations (Dootson et al., 2016;
Berry et al., 2010), which offer better ways of doing things for customers and improve
performance (Hult et al., 2004). Patsiotis et al. (2012) suggest that understanding the impact
of innovation on different categories of adopters and non-adopters is of potential value to
banks. Similarly, technology innovation focussed on customers is important for
organisations, because customers need to use the innovation to make it valuable to both
parties (Arts et al., 2011). However, there is limited study of customers’ perceptions on
innovation, and how it impacts DB experience. Baba (2012) notes that focussing on specific
innovation contributes more to performance (growth in market share) than adopting
different innovations at the same time. The relationship in innovation is tested, thus it is
hypothesized that:
H9. There is a positive relationship between DB innovation and customer experience.
Customer satisfaction, loyalty and experience. Different definitions of customer experience
have been given by Klaus and Maklan (2013) and Verhoef et al. (2009), while Liang et al.
(2009) suggest that customer satisfaction is overall customer experience. Studies in
customer experience and loyalty are limited, but what leads to customer satisfaction has
been tested in hospitality (Chi and Gursoy, 2009), internet banking (Amin, 2016; Raza et al.,
2015) and mobile banking ( Jun and Palacios, 2016) areas. Although these studies were
conducted in different countries, the antecedent of customer loyalty has predominantly been
tested using customer satisfaction, rather than customer experience. Klaus and Maklan’s
(2013) framework investigated customer experience, satisfaction and loyalty in high-contact
mortgage environment in the UK. This research tests the effect of DB experience, which is
contactless, through the hypotheses:
H10. There is a positive relationship between customer experience and satisfaction.
H11. There is a positive relationship between customer experience and loyalty.
Customer experience, satisfaction and loyalty and FP. Keisidou et al. (2013) investigate the
relationship between customer satisfaction and loyalty, and FP using ROA or ROI; NPM; and
ROE, while Anderson et al. (1994) used ROI. Chi and Gursoy (2009) asked hotel managers to
rate their FP in comparison to their competitors in terms of profitability, ROI and net profit.
Undoubtedly, different research aims call for different measures; some authors have
IJBM approximated profitability through loyalty. Reichheld et al. (2000) claim that improving
36,2 product quality enhances customer loyalty and profit through cross-buying,
recommendations and low-servicing cost, while Heskett et al. (2008) suggest that an
increase in customer satisfaction and loyalty can boost profitability. This argument has linked
customer loyalty to profit through customer lifetime value (CLV) – profit attributable to a
customer throughout their lifetime with a firm (Reichheld, 2003; Valenzuela et al., 2014). The
238 links between satisfaction and loyalty, and FP have been studied, and need to be tested in DB
experience, using financial ratios (ROE, cost-to-income ratio and NIM as FP1) and NPS value
(the loyalty effect on profit through CLV as FP2). The underpinning hypotheses are:
H12. There is a positive relationship between customer experience and FP.
H13. There is a positive relationship between customer satisfaction and FPFP.
H14. There is a positive relationship between customer loyalty and FP.
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

Customer satisfaction and loyalty. Limited customer loyalty studies have been conducted in
the DB sector. Researchers who study satisfaction and loyalty do not always consider
customer experience. For instance, the effect of service quality on customer satisfaction and
loyalty have been investigated (Levy and Hino, 2016; Kaura et al., 2015; Ladhari et al., 2011),
while Saleem et al. (2016) study the effect of customer loyalty, with customer satisfaction as
the moderator. There is a positive relationship between customer satisfaction and loyalty in
Greek banks (Keisidou et al., 2013) and in other banking studies (Fathollahzadeh et al., 2011;
Klaus and Maklan, 2013). These studies suggest that customer satisfaction can lead to
customer loyalty, which needs testing in DB. Therefore, we propose the hypothesis:
H15. There is a positive relationship between customer satisfaction and loyalty.

4. Methodology
This research used a web-based questionnaire method supported by e-mail (Ritter and
Sue, 2007; Kwak and Radler, 2002), and was conducted at the Sheffield Hallam University,
UK. A web-based tool called Bristol Online Survey (2015), developed at the Bristol
University for higher education research, was used to design the questionnaire. Prior to
sending out the main questionnaire, a pilot survey to assess validity and correct errors
was conducted. The questionnaire URL link was sent to ten selected respondents from
each of the sample strata to get their feedback and ensure the expected data would
address the research objectives. Their feedback was used to reshape the questionnaire,
remove ambiguity and make sure the questions were understandable. A total of 49
questions were asked after the pilot survey.
A total of 680 participants comprising 50 lecturers and 200 students from the Sheffield
Hallam University; 180 staff members from two large UK companies known to the
researchers; and 250 candidates from the researchers’ social media contacts (professional
LinkedIn) participated in the research study. The questionnaire’s URL was sent to the
selected respondents via e-mail and social media messenger in 2016. The survey was open
for three months. The sample profile comprised adults over the age of 18, living in the UK
and having an account with a major UK retail bank. The sample includes customers from
different backgrounds. The questionnaire asked the participants about the nature of
customer experience and the impact DB has on their lives. For example, whether they enjoy
it, find it convenient, what makes their experience with DB services good or bad, the quality
of DB received from their banks or whether it is reliable and accessible enough. Table I
shows the information on each measurement instrument.
All questions with the exception of NPS and customer profile were measured on a
5-point Likert scale and of the questionnaires administered, 206 usable questionnaires
Dimensions Definitions and literature evidence Items
UK customers’
perceptions
Perceived value (PV ) Saves money and time, usefulness, enjoyment, better deal online (e.g. Keisidou 4
et al., 2013; Liang et al., 2009; Garg et al., 2014; Fathollahzadeh et al., 2011; Chang
and Lin, 2015, Dootson et al., 2016)
Convenience (CONV ) Comfort, convenience, hassle-free (e.g. Keisidou et al., 2013; Knutson et al., 2007; 3
Karatepe et al., 2005; Garg et al., 2014; Klaus and Maklan, 2013; Jun and
Palacios, 2016; Wu, 2011) 239
Functional quality Interactive, easy to navigate, simple and intuitive (e.g. Keisidou et al., 2013; 3
(FQ) Garg et al., 2014; Monferrer-Tirado et al., 2016; Lee and Chung, 2009)
DB service quality Meeting and exceeding expectations, accessibility, reliability (e.g. Keisidou et al., 3
(DBSQ) 2013; Kaura et al., 2015; Levy and Hino, 2016; Parasuraman et al., 1988;
Ladhari et al., 2011; Amin, 2016)
Brand trust (BT) Staying loyal due to trustworthiness and brand (e.g. Keisidou et al., 2013; 3
Liang et al., 2009; Fathollahzadeh et al., 2011; Knutson et al., 2007; Akhter et al.,
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

2011; Levy and Hino, 2016)


Employee-customer Customer engagement, feedback, interactive support online, understanding 3
engagement (ECE) requirements (e.g. Karatepe and Aga, 2016; Verhoef et al., 2009; Garg et al., 2014;
Yee et al., 2010; Karatepe et al., 2005; Chi and Gursoy, 2009; Kanyurhi and
Akonkwa, 2016)
Perceived risk (PR) Security, cyber attack, fraud (e.g. Martins et al., 2014; Akinci et al., 2003; 3
Hanafizadeh et al., 2014; Jun and Palacios, 2016)
Perceived usability Ease of use, user-friendly, flexible and simple (e.g. Alalwan et al., 2016; Gu et al., 3
(PU) 2009; Klaus, 2013)
DB innovation (DBI) Better services, R&D, improving experience through innovation (e.g. Hult et al., 3
2004; Patsiotis et al., 2012; Dootson et al., 2016; Arts et al., 2011; Baba, 2012)
Customer experience Overall customer experience, meeting service needs and requirements 3
(CEQ) (e.g. Klaus and Maklan, 2013; Garg et al., 2014; Verhoef et al., 2009;
Liang et al., 2009)
Customer Overall satisfaction with interface, product and services (e.g. Fathollahzadeh 3
satisfaction (CSAT) et al., 2011; Keisidou et al., 2013; Klaus and Maklan, 2013; Amin, 2016;
Jun and Palacios, 2016)
Customer loyalty Staying longer, recommending friends and giving high NPS (e.g. Keisidou et al., 3
(CLY) 2013; Liang et al., 2009; Klaus and Maklan, 2013; Reichheld, 2003;
Levy and Hino, 2016; Amin, 2016)
Financial Financial ratios effect (e.g. Keisidou et al., 2013; Chi and Gursoy, 2009; 3
performance (FP1) Anderson et al., 1994) Table I.
Financial NPS effect via loyalty and CLV (e.g. Reichheld, 2003; Valenzuela et al., 2014; 3 Past studies on items
performance (FP2) Reichheld et al., 2000; Liang et al., 2009) for factor analysis

were returned giving a return rate of 30.29 per cent. The data were processed with SPSS
and SEM, AMOS software version 23. The customers were asked an 11-point Likert scale
question for measuring NPS and estimating CLV, recommended by Reichheld (2003).
The research selected six UK banks with public access to their financial status and
extracted three-year financial ratios from their financial reports. The NPS evidence from
customers was used to test for relationships between DB customer experience and the
banks’ FPFP based on their key financial ratios and NPS. For financial ratios, ROE, NIM
and cost-to-income ratio were used, with NPS of respondents from the six banks. All banks
used are regulated by the Financial Conduct Authority (FCA, 2015).

5. Measurement development
To construct the questionnaire items, the existing literature was studied extensively. This
was complemented with an “Idea Tournament” exercise, in which the research team
conducted a debate around the research objectives, to ensure that crucial aspects of the
IJBM study were covered. This ensured that enough questions were asked to measure the
36,2 study factors and support the hypotheses. Table I shows the items used to measure the
model factors.

5.1 Definition of items and content validity


Content validity ensures that the questionnaire items are valid and a thorough review of
240 the existing literature and a confirmatory pilot test involving another group of
respondents was carried out (Keisidou et al., 2013). This helps to improve framing of the
questionnaire items, so they are understood by different respondents. Factor analysis was
used to reduce questionnaire data into principal components that can produce the
information required for the study and testing the model hypotheses. Table I summarises
the dimensions, related literature, items and definitions of the 14 factors described in the
hypotheses development section.
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

6. Empirical analysis results


The results are presented in line with the research objectives as follows.

6.1 Respondents’ profile and frequency results


Objective 1: to provide exploratory data analyses of the study factors. Table II presents
descriptive analysis of the customer profile data obtained. It shows the results of some
important questions as percentages and frequencies of data distribution, approximated to
the nearest integer percentages as follows.
The results show a gender distribution of 70 per cent males and 30 per cent females,
which enable the results to accommodate male and female opinions. About 87 per cent of the
respondents fall into the 25-54 age group of active earners. About 96 per cent of the
respondents have college- and university-level education which makes them literate enough
to use DB effectively. The results also indicate that 88 per cent of participants are the
customers of one of the six major UK banks, which means that their responses relate to their
experience of UK DB.
Most of the customers have used DB for between 1 and 5 years (40.80 per cent), followed
by 6-10 years (37.90 per cent) and then 11 years and above (14.60 per cent). This indicates
that DB trend has gone up in the last ten years, showing the change in customers’
behaviour. Overall, about 93 per cent of the respondents have used DB at least one year and
15 per cent at least 11 years, which shows high uptake of DB noted in the literature. In terms
of customer loyalty, 28 per cent of the respondents have stayed loyal to their banks for one
to five years, 67 per cent for more than six years and 5 per cent for less than a year.
Conservatively, banks that significantly enhance their customers’ banking experiences can
attract less loyal customers, especially those 28 per cent in the first category.
In terms of DB channels usage, the order of prevalence is internet (51 per cent), mobile
(33 per cent) and telephone (15 per cent). Other types of DB accounted for 6 per cent. The use
of telephone banking therefore seems to be on the decline, while internet and mobile banking
are on the increase. Many customers use DB on a weekly basis (49.50 per cent), followed by
daily (34.00 per cent) and monthly (12.10 per cent). The result also shows that customers use
DB to carry out various services: check balance (16 per cent); fund transfer (15 per cent);
current account (14 per cent); pay bills (13 per cent); direct debit (12 per cent) and standing
order (10 per cent). A cross-tabulation between frequency of use, age groups, length of usage
and DB experience (suppressed in this paper) showed there are significant dependencies
among them; this means that bank marketing using these results can target the needs and
preferences of specific customer segments.
Measure Customer data Frequencies %
UK customers’
perceptions
Gender Male 145 70.40
Female 61 29.60
Age group 15-24 14 6.80
25-34 51 24.80
35-44 70 34.00
45-54 59 28.60 241
55-64 11 5.30
65+ 1 0.50
Educational level O Level/GCSE 8 3.90
College 34 16.50
University 164 79.60
Customers by bank Lloyds/Halifax 59 28.60
RBS/NatWest 29 14.10
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

HSBC 31 15.00
Barclays 39 18.90
Santander 18 8.70
Virgin Money 6 2.90
Others 24 11.70
Length of DB usage Less than 1 year 14 6.80
1-5 years 84 40.80
6-10 years 78 37.90
11+ years 30 14.60
Customers’ years of bank loyalty Less than 1 year 11 5.30
1-5 years 58 28.20
6-10 years 45 21.80
11+ years 92 44.70
Frequency of DB usage Daily 70 34.00
Weekly 102 49.50
Monthly 25 12.10
Less often 9 4.40
Most used DB channels Telephone banking 53 14.90
Internet banking 180 50.60
Mobile banking 117 32.90
Others 6 1.70
DB and Financial Services Savings 108 9.40
Check balance 181 15.80
Pay bills 147 12.80
Print statement 52 5.00
Transfer funds 172 15.00
Standing order 105 9.50
Current account 165 14.40
Buy insurance 27 2.40
Stock/shares 21 1.80
Direct debit 134 11.70
Mortgages 28 2.40
Others 5 0.40
NPS Detractors (0-6) 43 20.87
Passives (7-8) 85 41.26
Table II.
Promoters (9-10) 78 37.86 Bank customer profile
Overall rank NPS 16.99 and frequency
Note: n ¼ 206 information
IJBM Among the customers surveyed from the UK banks, NPS values consist of 21 per cent
36,2 detractors, 41 per cent passives and 38 per cent promoters. these scores were in response to
the question “Based on your experience with DB, how likely is it that you would recommend
your bank’s DB to a friend or colleague? Mark on a scale of 0 to 10?” These results show a
62 per cent potential for converting (digital) bank customers in the UK to the ultimate loyal
promoters, if individual banks pay attention to the customers’ needs, and devise appropriate
242 bank marketing strategies (digital and offline) for exceeding customers’ expectations.
Correlation analysis of the factors. To gauge the nature of relationships among pairs of
variables in the study, the 14 factors were cross-correlated as shown in Table III.
The results show strong positive correlations among the factors. For instance, there is a
significant positive relationship among “PV” and all the remaining variables, namely
“convenience”, “functional quality”, “DB service quality”, “brand trust”, “ECE”, “perceived
risk”, “perceived usability”, “DB innovation”, “customer satisfaction”, “customer loyalty”
and “customer experience”. All factors are significant.
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

Moreover, the four dependent variables in the study: customer experience quality (CEQ),
customer satisfaction (CSAT), customer loyalty (CLY) and FP measures (FP1 and FP2) show
that they are also strongly correlated with most of the other variables. These results mean
that the 14 factors are important mediators of customers’ perception of the value they derive
from DB, and its impact on FP. There is no significant relationship between “customer
experience” and “FP (FP2)”. This result may mean that the information on customer
experience is implicit in the strong positive correlations among customer experience,
satisfaction and loyalty (0.68 and 0.72, respectively). In sum, the above results partially
confirm most of the research hypotheses earlier developed, but this requires objective tests
of significance using factor and SEM analyses presented below.

6.2 Exploratory factor analysis (EFA) results


Objective 2: factor analyses of the data. EFA is performed to assess construct validity, with
regard to convergent and discriminant validity (Liang et al., 2009). EFA summarises
information from a group of variables into a smaller manageable number by allocating them
into distinct factors without significant loss of content and meaning (Hair et al., 1998). It is
performed with the method of principal component analysis and the varimax rotation for
extracting factors, which assumes factors are not related to each other. Minimum value
criteria for deleting items not meeting criteria are factor loadings (0.50) (Karatepe et al.,
2005), cross-loadings (0.40) or communalities (0.30) (Garg et al., 2014). For good factor
analysis and sampling adequacy, the value of Kaiser-Meyer-Olkin (KMO) statistic must be
at least 0.60 (Tabachnick and Fidell, 2001).
Unidimensionality, reliability, convergent and discriminant analyses. The two methods of
assessing unidimensionality of factors are EFA and confirmatory factor analysis (CFA).
The CFA is necessary where the structure model does not incorporate previously examined
the literature information (Keisidou et al., 2013; Sharma, 1996). EFA was performed. From
the 43 items used, 13 factors were produced. The cross-loadings and communalities W0.4.
The overall KMO is 0.866 and significant with po0.05, showing that factor analysis is
possible on the sample. The KMOs of all composite factors are all greater than 0.6. The
result justifies using EFA since all the criteria are met. Table IV shows the factor and
reliability analysis results along with total variance explained (TVE).
Columns 1-3 of Table IV show how the study factors are loaded on the questionnaire
items that significantly describe them with factor loadings greater than 0.50. This gives a
clue to the customer experiences and expectations, which determine the factors.
For example, the “PV” variable is significantly associated with Qs 3, 6, 8 and 36 in the
questionnaire, namely “saves money”, “saves time”, “usefulness”, “enjoyment”
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

Mean SD PV Conv FQ DBSQ BT ECE PR PU DBI CSAT CLY CEQ FP1 FP2

17.49 2.30 PV
14.12 1.58 Conv 0.768**
12.27 1.98 FQ 0.704** 0.630**
11.55 2.11 DBSQ 0.651** 0.558** 0.720**
11.00 1.94 BT 0.467** 0.324** 0.469** 0.617**
11.93 1.73 ECE 0.509** 0.401** 0.421** 0.400** 0.430**
10.41 2.24 PR 0.471** 0.349** 0.520** 0.641** 0.554** 0.351**
12.54 2.00 PU 0.706** 0.596** 0.802** 0.793** 0.488** 0.360** 0.493**
11.99 1.99 DBI 0.660** 0.535** 0.702** 0.666** 0.499** 0.451** 0.481** 0.668**
12.15 1.97 CSAT 0.685** 0.551** 0.806** 0.804** 0.520** 0.351** 0.492** 0.847** 0.679**
12.05 2.20 CLY 0.698** 0.593** 0.730** 0.783** 0.581** 0.481** 0.479** 0.783** 0.703** 0.761**
11.59 2.12 CEQ 0.663** 0.538** 0.715** 0.659** 0.504** 0.564** 0.427** 0.713** 0.597** 0.682** 0.718**
192.96 78.70 FP1 0.298** 0.311** 0.228** 0.214** 0.149* 0.205** 0.145* 0.211** 0.341** 0.169* 0.262** 0.230**
18.42 14.42 FP2 0.174* 0.192** 0.227** 0.296** 0.167* 0.089 0.251** 0.216** 0.247** 0.212** 0.239** 0.130 0.570**
Notes: n ¼ 206. *p o0.05, **p o0.01, significance (two-tailed)
perceptions

243

Correlation analysis of
UK customers’

the factors
Table III.
IJBM Factors Items Loadings KMO TVE Bartlett’s test significance Commualities Cronbach’s α
36,2
PV Q3 0.774 0.744 57.888 0.000 0.714 0.736
Q6 0.584 0.617
Q8 0.568 0.602
Q36 0.538 0.631
CONV Q4 0.790 0.732 82.423 0.000 0.791 0.893
244 Q5 0.780 0.802
Q7 0.680 0.699
FQ Q11 0.590 0.689 70.140 0.000 0.588 0.786
Q19: 0.650 0.678
Q26 0.684 0.665
DBSQ Q21 0.748 0.690 66.760 0.000 0.777 0.749
Q25 0.576 0.670
Q27 0.599 0.663
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

BT Q32 0.550 0.615 54.400 0.000 0.504 0.754


Q33 0.718 0.686
Q34 0.640 0.696
ECE Q12 0.586 0.606 50.799 0.000 0.628 0.706
Q13 0.768 0.659
Q31 0.501 0.582
PR Q28 0.706 0.648 70.577 0.000 0.647 0.778
Q29: 0.788 0.744
Q30 0.739 0.747
PU Q9 0.767 0.691 72.486 0.000 0.770 0.803
Q17 0.736 0.721
Q22 0.668 0.704
DBI Q10 0.727 0.622 62.712 0.000 0.764 0.712
Q37 0.783 0.745
Q38 0.598 0.735
CEQ Q14 0.771 0.667 66.022 0.000 0.699 0.736
Q15 0.568 0.591
Q16 0.680 0.589
CSAT Q18 0.671 0.714 76.267 0.000 0.668 0.844
Q20 0.783 0.774
Q35 0.765 0.670
CLY Q23 0.797 0.689 64.362 0.000 0.767 0.710
Q24 0.616 0.595
Q39 0.553 0.593
FP1 ROE 0.795 0.704 81.025 0.000 0.710 0.882
NIM 0.893 0.861
Cost-to-income 0.863 0.839
ratio
Table IV. FP2 Q44: NPS
Exploratory factor Detractors 0.718 0.690 71.853 0.000 0.668 0.764
and reliability Passives 0.859 0.818
analysis results Promoters 0.742 0.649

and “better deal online”. Table I summarises these descriptors as useful evidence for bank
marketing strategies.
Reliability item analysis refers to the internal consistency of the factors (Chu and
Murramann, 2006), which is measured using Cronbach’s coefficient α (Fornell and Larcker,
1981; Churchill, 1979). For all factors, coefficient α was computed and all values range from
0.706 to 0.893. The values of α exceed the minimum 0.7 score (Nunnally, 1978) and 0.6 reported
in Garg et al. (2014). The result shows a construct reliability which indicates internal
consistency. Therefore, improving the value of α for each cluster of items is not required.
The results show the unidimensionality of the measures, as each item is related to only one UK customers’
fundamental construct (Garg et al., 2014; Gerbing and Anderson, 1988). perceptions
Convergent validity is the degree to which several methods of measuring a factor
provide the same output (Keisidou et al., 2013). The acceptable value of convergent validity
is 0.5 for all items loading, while Garg et al. (2014) added that all items should load to only
one factor with an eigenvalue W1. All items loaded to their predestined factor with an
eigenvalue W1. In Table IV, all items bear loadingsW0.5, which complete the criteria for 245
convergent validity. Convergent validity is also examined by identifying whether the
maximum likelihood loading of each indicator is significant for its underlying construct
(Peter, 1981). The results also show that all loadings W0.5, KMO is significant ( p o0.05) and
TVE values ranges from 50.80 to 82.42. This means that more than half of the variance is
extracted. All items in Table IV are significant ( po0.05). The NPS detractors, passives and
promoters range from 15.29 to 17.49, 34.63 to 38.73 and 32.68 to 36.93 respectively for the six
banks. The TVE values, factor loadings above 0.5, coupled with the range of t-values also
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

prove convergence of factor items (Garg et al., 2014). All these indicate good convergence
validity between the items within the instruments.
The discriminant validity is about dissimilar constructs and items used in factor analysis
being different (Keisidou et al., 2013). Table III was also constructed for meeting these
criteria, so the correlation coefficients of the factors along the diagonal are compared with
Cronbach’s α values. Churchill (1979) and Keisidou et al. (2013) state that Cronbach’s values
of the factors should be higher than the correlation values, indicating that the correlation
among the factors is lower than Cronbach’s α. The correlation values are below the
maximum Cronbach’s α of 0.893. The result confirms the discriminant validity criterion.

6.3 SEM
Objective 3: SEM of factors links and test of subsidiary and primary hypotheses. The factor-
related data were analysed using multivariate techniques (e.g. correlation and SEM
analyses). They have the ability to simultaneously examine a number of dependent linear
relations, where one or more constructs (variables) are both dependent and independent
(Hair et al., 1998). Most of the variables depend on each other as shown in the correlation
analyses, for which reason each can serve as a predictor of the other.
SEM technique was used to examine the model factors, as it has the ability to test and draw
relationships on the paths of a model. The most commonly used model fit measures in SEM are:
χ2/degree of freedom (χ2/df), goodness of fit index (GFI), comparative fit index (CFI),
Tucker-Lewis index (TLI), normed fit index (NFI), incremental fit index (IFI) and root mean
square error of approximation (RMSEA) (Keisidou et al., 2013; Garg et al., 2014). Table V
presents the overall fit values of the model obtained by examining the causal relationships
among the factor variables, which indicates a moderately good fit for both PF1 and PF2.
All demonstrated good fit except NFIo0.9. Table V shows the SEM parameters for the model
fit for FP1 and FP2.

Overall fit of the model


Parameters PF1 value PF2 value

χ /df
2
2.11 2.09
CFI 0.907 0.911 Table V.
TLI 0.901 0.906 Model fit for financial
NFI 0.862 0.864 performance
RMSEA 0.068 0.065 FP1 and FP2
IJBM Table V shows the path loadings for the SEM model fit for FP1 (χ2/df) ¼ 2.11, p ¼ 0.00,
36,2 CFI ¼ 0.907, TLI ¼ 0.901, NFI ¼ 0.862 and RMSEA ¼ 0.068. The model fit for FP2
(χ2/df ) ¼ 2.09, p ¼ 0.00, CFI ¼ 0.911, TLI ¼ 0.906, NFI ¼ 0.864 and RMSEA ¼ 0.065.
Figure 2 shows SEM of the factors using financial ratios as FP1 indicators, along with
path coefficients.
Figure 3 shows the SEM of the factors using NPS (CLV ) as FP2 indicators.
246 Table VI shows the SEM test results for all the factors against customer experience as
the dependent variable while other factors are independent variables for FP1 and FP2.
There are two types of FP tests in the Model, hence (a) and (b). H ¼ Hypothesis.
Table VI shows, for instance, there is a significant positive relationship between “PV”,
“functional quality”, “DB service quality”, “ECE”, “perceived risk”, “perceived usability” and
“customer experience”. “Convenience”, “brand trust” and “DB innovation” were not significant
predictors. There is a significant positive relationship between “customer loyalty” and both
“FP1” and “FP2”, respectively. These types of relationships generate the list of accepted and
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

rejected hypotheses. These accepted hypotheses are the significant results that should inform
bank marketing strategies. For example, to improve customer experience, banks should
consider the factors listed above, especially the sets of accepted hypotheses on factors which
influence “customer experience”. This applies to other hypotheses in the table.

6.4 Customer experience factors against customer demographics


Objective 4: ANOVA tests on study factors and customer characteristics. The section applies
a one-way ANOVA test to identify the relationship between respondents’ profile variables
and 12 factors in the model. These tests help to identify how the factors are affected by
different bank types and customer profiles. Table VII shows the results.

Perceived value 0.14

em1

Convenience
0.40
–0.05
Customer
satisfaction
Functional quality 0.31 em4 –0.10
0.63 em3
0.32
0.72
Digital BK 0.12 0.55
service quality 0.07
0.10 0.51
Financial performance (ratios)
0.36 Customer
0.09 experience
0.35 Brand/trust
em2 0.25
0.71
0.30 0.35

Employee 0.60
customer Egt

–0.10
Customer loyalty

Perceived risk

0.31
0.47 0.12
Perceived usability

Figure 2.
Model of the factors –0.03
using FP1 Digital banking
innovation
Perceived value 0.14
UK customers’
perceptions
em1

Convenience
0.40
–0.05
Customer
satisfaction
Functional quality
0.31 em4 0.10
247
0.32 0.63 em3
0.72

Digital BK 0.12
service quality 0.55 0.07
–0.10 0.51
Financial performance (CLV)
0.36 Customer
0.09 experience
0.35 Brand/trust
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

em2
0.71 0.35 0.22
0.30

Employee
0.60
customer Egt

–0.10
Customer loyalty

Perceived risk

0.31
0.47 0.12
Perceived usability

Figure 3.
–0.03 Model of the factors
Digital banking using FP2
innovation

For PV, for example, Table VII shows that it is significantly affected by customers’ age, type
of bank, frequency of DB usage and NPS value. All the factors are generally affected by
most customer characteristics, apart from educational level and gender. This shows that all
the influencing customer characteristics are potential inputs into bank marketing strategies
aimed at influencing customers’ perceptions of the study factors, with the factors linked to
accepted hypotheses above probably more important to focus on.

7. Theoretical and managerial implications of results


This section summarises the results in light of the research objectives and implications for
bank marketing. The results in Objective 1 show that DB experience in the UK differs by
customer characteristics; hence bank marketing strategies aim to satisfy the customers
need by targeting specific segments. Objective 2 results show the different questionnaire
items and underpinning customer expectations, which moderate DB variables, thereby
providing further evidence base for constructing appropriate bank marketing strategies
referred to above. Objective 3 results provide the sets of positive hypotheses
linking customer experience to the first nine factors, and importantly customer
experience and other dependent variables in the study: customer satisfaction, loyalty
and FP (FP1 and FP2). New SEM path analyses generated additional hypotheses linking
some of the explanatory factors to each other. These provide handles on plausible bank
marketing strategies to consider in order to enhance customers’ DB experience. Finally,
Objective 4 ANOVA test results show how different DB experience factors are affected by
different customer characteristics. This reinforces the evidence base for future bank
marketing approaches suggested above.
IJBM Research
36,2 hypotheses Path coefficient Remark

H1 Perceived value has a positive relationship with customer experience 0.14* Accept
H2 Convenience has a positive relationship with customer experience −0.05 Reject
H3 Functional quality has a positive relationship with customer experience 0.31** Accept
H4 DB service quality has a positive relationship with customer experience 0.12* Accept
248 H5 Brand trust has a positive relationship with customer experience 0.09 Reject
H6 Employee customer engagement has a positive relationship with 0.30** Accept
customer experience
H7 Perceived risk has a positive relationship with customer experience −0.10* Accept
H8 Perceived usability has a positive relationship with customer experience 0.31** Accept
H9 DB Innovation has a positive relationship with customer experience −0.03 Reject
H10 Customer experience has a positive relationship with customer satisfaction 0.63** Accept
H11 Customer experience has a positive relationship with customer loyalty 0.35** Accept
H15 Customer satisfaction has a positive relationship with customer loyalty 0.51** Accept
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

H13a Customer satisfaction has a positive relationship with FP1 −0.10 Reject
H14a Customer loyalty has a positive relationship with FP1 0.25* Accept
H12a Customer experience has a positive relationship FP1 0.10 Reject
H13b Customer satisfaction has a positive relationship FP2 0.10 Reject
H14b Customer Loyalty has a positive relationship with FP2 0.22* Accept
H12b Customer experience has a positive relationship FP2 −0.10 Reject
New SEM path
Functional quality has a positive relationship with DB service quality 0.72** Accept
Employee customer engagement has a positive relationship with functional 0.36** Accept
quality
DB innovation has a positive relationship with perceived risk 0.47** Accept
Employee customer engagement has a positive relationship with DB 0.35** Accept
service quality
Table VI. Brand trust has a positive relationship with convenience 0.32** Accept
Hypothesis test Perceived value has a positive relationship with perceived usability 0.77** Accept
results for DB innovation has a positive relationship with employee customer engagement 0.12* Accept
FP1 and FP2 Notes: n ¼ 206. *p o0.05; **p o0.01

Age Educational Frequency of Length of DB Length of bank NPS


group level Gender Banks DB usage usage loyalty value
Factors F-value F-value F-value F-value F-value F-value F-value F-value

PV 5.41** 0.49 0.09 5.22** 35.95** 12.29** 1.99 39.68**


Conv 2.88** 0.03 0.18 5.33** 30.98** 12.08** 0.51 22.02**
FQ 2.77** 0.19 0.01 2.73** 25.73** 6.85** 0.65 42.72**
DBSQ 3.27** 0.72 0.69 4.84** 18.82** 6.81** 0.78 54.02**
BT 2.98** 1.73 1.10 1.57 8.03** 4.56** 2.54** 19.19**
ECE 1.62 2.22 0.56 2.27* 5.13** 2.15 3.00** 7.22
PR 0.64 0.58 0.55 3.19** 8.33** 1.29 1.33 13.32**
PU 6.48** 0.66 0.02 2.47* 19.98** 11.62** 0.23 44.83**
DBI 2.47* 0.35 0.03 4.78** 27.81** 4.28** 0.34 48.26
Table VII. CSAT 3.71** 0.64 0.29 2.09* 18.19** 11.85** 0.65 82.98**
Analysis of variance CLY 5.04** 1.47 0.03 3.31** 31.20** 11.17** 1.70 80.29**
between factors and CEQ 4.17** 2.15 0.04 2.86** 19.03** 13.14** 0.33 35.29**
customer data Notes: n ¼ 206. *p o0.05; **p o0.01

Theoretically, the research produces FP1 and FP2 from SEM and factor-based models which
will support further research in DB, customer experience and FP, in the UK, given that no
such studies were previously done along these lines, with few outside the UK (Keisidou et al.,
2013; Amin, 2016; Kaura et al., 2015; Jun and Palacios, 2016). Specifically, while there are other
studies in contact service marketing from different perspectives and countries (Liang et al., UK customers’
2009; Ladhari et al., 2011), and customer experience (Klaus and Maklan, 2013), the results in perceptions
this paper provide UK-based influences on DB customer experience through the relevant
hypotheses. Furthermore, the research established the factors that affect “customer
experience”, namely “functional quality”, “ECE”, “service quality”, “perceived usability”,
“functional quality” (e.g. better interfaces), “perceived risk” (e.g. security), “PV” (e.g. Being
useful, cost savings) and the “Perceived Usability” (e.g. ease of use) of DB. The managerial 249
implication is that to improve DB experience, banks should pay attention to these factors,
while theoretically they can serve as building blocks for further research.
The research established links between these factors and customer profile data, for
example, “PV”, “perceived usability” and “convenience” have relationships with customers’
“frequency of DB usage”, showing the three factors can determine whether customers use
DB very often or not. Full results are demonstrated in Table VII. This result will help banks
in customer acquisition and retention, and strategic marketing of products, so certain
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

customer segments can be targeted, based on the factors that are significant to them.
There are contextual similarities and differences between some findings elsewhere and
this study, which show the need for location-specific studies of DB in support of bank
marketing and FP. For instance, Garg et al.’s (2014) results from Indian banks showed that
gender, marital status, age, education level and income have significant relationships with
some factors of customer experience, but this research result shows that customers’ DB
experience in the UK does not depend on educational level and gender.
Result in “PV” is consistent with the service marketing theory and corresponds to
Dootson et al.’s (2016) finding that expected value draws customers towards performing an
action. Similar results were found in Greek banks (Keisidou et al., 2013), and in e-commerce
marketing in India (Piyathasanan et al., 2015) and Korea (Ko et al., 2009). This shows that
generally customers are looking for value and therefore managers should be mindful of this.
“Convenience” positively affects customer satisfaction (Keisidou et al., 2013; Jun and
Palacios, 2016) and customer experience (Garg et al., 2014; Wu, 2011) in locational activities.
This research finding differs from Jun and Palacios (2016) and Garg et al. (2014), but coheres
with Keisidou et al.’s (2013) finding. There is no consensus among the authors on
convenience; it may be associated more with acceptance and location than operation of DB,
as customers can access it from anywhere.
“Functional quality” affects UK customers’ DB experience, and customer satisfaction and
loyalty in Spanish (Monferrer-Tirado et al., 2016) and Greek (Keisidou et al., 2013) banks by
incorporating offline activities. Studies on “service quality” effect on customer experience
are limited except ones conducted in customer satisfaction and loyalty (Kaura et al., 2015;
Levy and Hino, 2016), and in contact services (Keisidou et al., 2013). Previous studies showed
that service quality affects customer satisfaction and loyalty on internet banking in Saudi
Arabia (Amin, 2016) and Pakistan (Raza et al., 2015), and on mobile banking in the USA
( Jun and Palacios, 2016). Meanwhile, this research highlighted that service quality affects
customer experience, and consequently leads to customer satisfaction and loyalty. These
results offer further theoretical and marketing insights across countries in DB.
Bank employees’ attributes were found to be an important link in customer service
delivery in non-DB environments (Karatepe and Aga, 2016; Karatepe et al., 2005; Verhoef
et al., 2009; Yee et al., 2010). In this research, ECE influences their ability to design DB that
improves customer experience, hence highlighting the relevance of customer feedback in
influencing positive customer behaviour in DB. “Perceived Risk” result showed a negative
impact on DB experience. Jun and Palacios (2016) also found security to affect service
quality of mobile banking. Perceived risk affects customers’ DB behaviour, and should be
minimised through enhanced security. “Brand trust” affects customer choices and improves
customer satisfaction and loyalty (Liang et al., 2009; Levy and Hino, 2016), suggesting that
IJBM brand relates to satisfaction and loyalty more than DB experience, which is about
36,2 customers’ perceptions within the application interfaces. Therefore, “Brand trust” and
customer experience need to be explored further.
“Perceived usability” affects telebanking experience in Jordanian banks (Alalwan et al.,
2016) and e-commerce experience (Klaus, 2013). Similarly, in UK DB sector, this factor
affects bank customer experience, which extends knowledge in the area. Banks benefit from
250 interactive service innovations (Dootson et al., 2016), but findings on “DB innovation”
through customers’ perceptions contradict this. Although innovation is important in service
development, customers are more interested in the benefits than the innovation itself.
It suggests that DB innovation should focus on the PV customers derive from innovation,
and confirms Patsiotis et al.’s (2012) study which suggests that understanding the impact of
innovation on customers is of potential benefit to banks.
“Customer experience” is positively related to “satisfaction” and “loyalty”, as well as
“satisfaction” being related to “loyalty”. The result between customer satisfaction and
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

loyalty alone relates to Jun and Palacios’ (2016) finding on mobile banking study in the USA,
and Amin (2016) and Raza et al.’s (2015) studies of internet banking in Saudi Arabia and
Pakistan, hence showing similarity of customers across countries in terms of customer
satisfaction and loyalty. Klaus and Maklan (2013) found similar results in UK high-contact
services using questionnaires distributed to mortgage and luxury goods customers, but this
research demonstrated customer experience, satisfaction and loyalty moderators in DB,
which extends the theory.
The managerial implications of the results hinge on the above customer experience
factors. Bank managers will know the factors that make a customer accept one bank’s DB
over another, which should be considered when implementing DB. Improving these factors
can help capture and retain customers; making them accept DB and stay loyal, leading to FP
improvement. PVs (e.g. cost saving, better deals, online interaction, enjoyment and time
saving) play a crucial role. To improve customer experience, banks should offer value-added
services, improve service quality, functional quality and security. Bank employees should
constantly engage with customers through feedback to be attuned to their requirements.
More customers access services through internet banking than other channels, and
mobile banking demand is on the increase while telephone banking demand is declining.
This emerging trend indicates that managers should invest in and focus more on mobile
banking services. Therefore, as more mobile banking technologies emerge, banks will have
to balance customer needs with design and security issues, and ensure that different
customers’ needs are fulfilled to improve loyalty. For major services offered through DB
channels, checking balance accounted for the highest transaction, followed by funds
transfer, as illustrated in Table II. This helps banks to know the digital channels to focus on
and value-added services to provide, helping them in strategic service marketing.
Methodologically, the research uses ROE, cost-to-income ratio and NIM as indicators on
FP1, and uses NPS loyalty effect of CLV on FP2. The results showed a significant positive
relationship between “customer loyalty” and “FP” on both FP1 and FP2. This indicates that
banks can improve FP through offering good DB experience, which improves loyalty. Loyal
customers pay premiums, recommend friends, and require less service costs and effort to
retain. Keisidou et al. (2013) used ROI/ROA, NPM and ROE to test FP. While NPM is good, the
measure is not consistent when banks report in different currencies. This research used
financial indicators reported in percentages, which are consistent across the six banks.
Cost-to-Income ratio was also used due to the impact DB can make on efficiency savings. Most
studies investigated FP using financial ratios (Keisidou et al., 2013; Chi and Gursoy, 2009).
This research also tested NPS effect on banks’ performance in DB, hence contributing to
studies that have attributed loyalty to FP (Reichheld, 2003; Valenzuela et al., 2014; Liang et al.,
2009), and offering theoretical link between customer relationship and bank marketing. Unlike
this study, some studies stopped at customer experience, satisfaction and loyalty, excluding UK customers’
FPFP (e.g. Klaus and Maklan, 2013; Garg et al., 2014; Jun and Palacios, 2016). The research offers perceptions
a broader linkage of phenomena in DB experience which can serve future study.
The research showed that “frequency of use”, “age group” and “length of usage” positively
affect DB experience. This implies that customers who use DB frequently are the ones enjoying
it, having a good experience and using it for a long time. The research showed that the uptake
of DB has improved in the last few years due to benefits to both banks and customers. 251
More customers are using DB than going to the branches, and banks are closing branches as a
result (BBC, 2016; French et al., 2013). That said, banks should consider why some customers
do not use DB frequently and some of the factors that affect customers have been highlighted.

8. Conclusion
This paper presented an integrated understanding of customers’ perceptions of the links
among DB, customer experience, satisfaction, loyalty, two measures of FP, namely financial
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

ratios (FP1) and NPS criteria (FP2), and the implications of these links for bank marketing.
The research demonstrated that banks can improve FP using DB. The overall customers’ NPS
is positive (16.99); however, passive customers are more than promoters. UK banks need to
target passive customers and turn them into promoters to improve their NPS using the
identified factors, which can help improve customer experience and FP. Customers are looking
for value and demanding more mobile banking services, so banks should be delivering these.
The methods and the type of analysis undertaken show the robustness of the developed DB
models, which can be used to explore customer experience and FP in future studies.

8.1 Limitations
This study provides thorough theoretical foundations and robust empirical results;
however, it is not free from limitations. There are general issues with the questionnaire
research such as a low response rate (Ritter and Sue, 2007). The web-based approach has
enabled the research to maintain anonymity and prevent respondents submitting
incomplete questionnaires, which is an advantage. There was a 30.29 per cent response rate,
which is not unexpected for web-based questionnaires (Kwak and Radler, 2002). The
financial data used were taken from the banks’ annual reports. Quite often information
reported in them is targeted towards shareholders; however previous researchers have
found them useful. The research concentrates on UK bank customers.

8.2 Future research


The research needs to be replicated in banks and extended to other countries, for example
developing countries in Africa. Africa is one of the up-and-coming continents where mobile
payments and DB are beneficial due to the large population living in remote areas. Further
research is needed to understand whether there are other factors that affect bank’s customer
experience and FP in those contexts. Extending the research to specific banks and bank
employees’ perceptions will triangulate the results with those from customers’ perceptions.
Covering all these additional lines of research will help to develop more robust digital bank
marketing theory in future.

References
Akhter, W., Abbasi, A.S., Ali, I. and Afzal, H. (2011), “Factors affecting customer loyalty in Pakistan”,
African Journal of Business Management, Vol. 5 No. 4, pp. 1167-1174.
Akinci, S., Aksoy, S. and Atilgan, E. (2003), “Adoption of internet banking among sophisticated
consumer segments in an advanced developing country”, The International Journal of Bank
Marketing, Vol. 22 No. 3, pp. 212-232.
IJBM Alalwan, A.A., Dwivedi, Y.K., Rana, N.P. and Simintiras, A.C. (2016), “Jordanian consumers’ adoption of
36,2 telebanking influence of perceived usefulness, trust and self-efficacy”, International Journal of
Bank Marketing, Vol. 34 No. 5, pp. 690-709.
Al-Eisa, A.S. and Alhemoud, A.M. (2009), “Using a multiple-attribute approach for measuring customer
satisfaction with retail banking services in Kuwait”, The International Journal of Bank
Marketing, Vol. 27 No. 4, pp. 294-314.
Amin, M. (2016), “Internet banking service quality and its implication on e-customer satisfaction and
252 e-customer loyalty”, International Journal of Bank Marketing, Vol. 34 No. 3, pp. 280-306.
Andaleeb, S.S., Rashid, M. and Rahman, Q.A. (2016), “A model of customer-centric banking practices
for corporate clients in Bangladesh”, International Journal of Bank Marketing, Vol. 34 No. 4,
pp. 458-475.
Anderson, E.W., Fornell, C. and Lehmann, D.R. (1994), “Customer satisfaction, market share, and
profitability: findings from Sweden”, Journal of Marketing, Vol. 58 No. 3, pp. 53-66.
Arts, J.W.C., Frambach, R.T. and Bikmolt, T.H.A. (2011), “Generalizations on consumer innovation
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

adoption: a meta-analysis on drivers of intention and behavior”, International Journal of


Research in Marketing, Vol. 28 No. 2, pp. 134-144.
Baba, Y. (2012), “Adopting a specific innovation type versus composition of different innovation types
case study of a Ghanaian bank”, International Journal of Bank Marketing, Vol. 30 No. 3, pp. 218-240.
BBC (2016), “Banks close more than 600 branches over the past year”, available at: www.bbc.com/
news/business-36268324 (accessed 16 June 2016).
Berry, L.L., Bolton, R.N., Bridges, C.H., Meyer, J., Parasuraman, A. and Seiders, K. (2010),
“Opportunities for innovation in the delivery of interactive retail services”, Journal of Interactive
Marketing, Vol. 24 No. 2, pp. 155-167.
Berry, L.L., Carbone, L.P. and Haeckel, S.H. (2002), “Managing the total customer experience”,
MIT Sloan Management Review, Vol. 43 No. 3, pp. 85-89.
Brakus, J.J., Schmitt, B.H. and Zarantonello, L. (2009), “Brand experience: what is it? How is it
measured? Does it affect loyalty?”, Journal of Marketing, Vol. 73 No. 3, pp. 52-68.
Bristol Online Survey (2015), “The online survey tool designed for academic research”, available at:
www.onlinesurveys.ac.uk/ (accessed 16 November 2015).
Chang, S.-H. and Lin, R. (2015), “Building a total customer experience model: applications for the travel
experiences in Taiwan’s creative life industry”, Journal of Travel & Tourism Marketing, Vol. 32
No. 4, pp. 438-453.
Chi, C.G. and Gursoy, D. (2009), “Employee satisfaction, customer satisfaction and financial
performance: an empirical examination”, International Journal of Hospitality Management,
Vol. 28 No. 2, pp. 245-253.
Chu, K.H. and Murramann, S.K. (2006), “Development and validation of the hospitality emotional labor
scale”, Tourism Management, Vol. 27 No. 6, pp. 1181-1191.
Churchill, G.A. (1979), “A paradigm for developing better measures of marketing constructs”, Journal
of Marketing Research, Vol. 16 No. 1, pp. 64-73.
Cortiñas, M., Chocarro, R. and Villanueva, M.L. (2010), “Understanding multi-channel banking
customers”, Journal of Business Research, Vol. 63 No. 11, pp. 1215-1221.
Daniel, E. (1999), “Provision of electronic banking in the UK and the Republic of Ireland”, International
Journal of Bank Marketing, Vol. 17 No. 2, pp. 72-82.
Davis, F.D. (1989), “Perceived usefulness, perceived ease of use, and user acceptance of information
technologies”, MIS Quarterly, Vol. 13 No. 3, pp. 319-340.
Dootson, P., Beatson, A. and Drennan, J. (2016), “Financial institutions using social media – do
consumers perceive value?”, International Journal of Bank Marketing, Vol. 34 No. 1, pp. 9-36.
Fathollahzadeh, M., Hashemi, A. and Kahreh, M.S. (2011), “Designing a new model for determining
customer value satisfaction and loyalty towards banking sector of Iran”, European Journal of
Economics, Finance and Administrative Sciences, Vol. 28 No. 1, pp. 126-138.
FCA (2015), “Financial Conduct Authority”, available at: www.fca.org.uk/ (accessed 16 July 2015). UK customers’
Fornell, C. and Larcker, D. (1981), “Evaluating structural equation models with unobservable variables perceptions
and measurement error”, Journal of Marketing Research, Vol. 18 No. 1, pp. 39-50.
French, S., Leyshon, A. and Meek, S. (2013), The Changing Geography of British Bank and Building
Society Branch Networks, 2003-2012, University of Nottingham, Nottingham.
Garg, R., Rahman, Z. and Qureshi, M.N. (2014), “Measuring customer experience in banks: scale
development and validation”, Journal of Modelling in Management, Vol. 9 No. 1, pp. 87-117. 253
Gerbing, D.W. and Anderson, J.C. (1988), “An updated paradigm for scale development incorporating
unidimensionality and its assessment”, Journal of Marketing Research, Vol. 25 No. 2, pp. 186-192.
Grace, D. and O’Cass, A. (2004), “Examining service experiences and post-consumption evaluations”,
Journal of Services Marketing, Vol. 18 No. 6, pp. 450-461.
Grönroos, C. (1982), “An applied service marketing theory”, European Journal of Marketing, Vol. 16
No. 7, pp. 30-41.
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

Grönroos, C. (1984), “A service quality model and its marketing implications”, European Journal of
Marketing, Vol. 18 No. 4, pp. 36-44.
Gu, J.-C., Lee, S.-C. and Suh, Y.-H. (2009), “Determinants of behavioral intention to mobile banking”,
The Journal of Expert Systems with Applications, Vol. 36 No. 9, pp. 11605-11616.
Hair, J.F., Anderson, R.E., Tatham, R.L. and Black, W.C. (1998), Multivariate Data Analysis, 5th ed.,
Prentice-Hall, Englewood Cliffs, NJ.
Hanafizadeh, P., Behboudi, M., Koshksaray, A.A. and Tabar, M.J.S. (2014), “Mobile-banking adoption
by Iranian bank clients”, Journal of Telematics and Informatics, Vol. 31 No. 1, pp. 62-78.
Heskett, J.L., Jones, T.O., Loveman, G.W., Sasser, W.E. Jr and Schlesinger, L. (2008), “Putting the
service-profit chain to work”, Harvard Business Review, Vol. 86 Nos 7/8, pp. 118-129.
Hoehle, H., Scornavacca, E. and Huff, S. (2012), “Three decades of research on consumer adoption and
utilization of electronic banking channels: a literature analysis”, Decision Support Systems,
Vol. 54 No. 1, pp. 122-132.
Holbrook, M.B. (2000), “The millennial consumer in the texts of our times: experience and
entertainment”, Journal of Macromarketing, Vol. 20 No. 2, pp. 178-192.
Hult, G.T., Hurley, R.F. and Knight, G.A. (2004), “Innovativeness: its antecedents and impact on
business performance”, Industrial Marketing Management, Vol. 33 No. 5, pp. 429-438.
Jun, M. and Palacios, S. (2016), “Examining the key dimensions of mobile banking service quality: an
exploratory study”, International Journal of Bank Marketing, Vol. 34 No. 3, pp. 307-326.
Kanyurhi, E.B. and Akonkwa, D.B.M. (2016), “Internal marketing, employee job satisfaction, and
perceived organizational performance in microfinance institutions”, International Journal of
Bank Marketing, Vol. 34 No. 5, pp. 773-796.
Karatepe, O.M. and Aga, M. (2016), “The effects of organization mission fulfillment and perceived
organizational support on job performance: the mediating role of work engagement”,
International Journal of Bank Marketing, Vol. 34 No. 3, pp. 368-387.
Karatepe, O.M., Yavas, U. and Babakus, E. (2005), “Measuring service quality of banks: scale development
and validation”, Journal of Retailing and Consumer Services, Vol. 12 No. 5, pp. 373-383.
Kaura, V., Durga Prasad, C.S. and Sharma, S. (2015), “Service quality, service convenience, price and
fairness, customer loyalty, and the mediating role of customer satisfaction”, International
Journal of Bank Marketing, Vol. 33 No. 4, pp. 404-422.
Keisidou, E., Sarigiannidis, L., Maditinos, D.I. and Thalassinos, E.I. (2013), “Customer satisfaction,
loyalty and financial performance”, International Journal of Bank Marketing, Vol. 31 No. 4,
pp. 259-288.
Kim, S.H., Cha, J.M., Knutson, B.J. and Beck, J.A. (2011), “Development and testing of the Consumer
Experience Index (CEI)”, Managing Service Quality: An International Journal, Vol. 21 No. 2,
pp. 112-132.
IJBM Klaus, P. (2013), “The case of Amazon.com: towards a conceptual framework of online customer service
36,2 experience (OCSE) using the emerging consensus technique (ECT)”, Journal of Services
Marketing, Vol. 27 No. 6, pp. 443-457.
Klaus, P. and Maklan, S. (2013), “Towards a better measure of customer experience”, International
Journal of Market Research, Vol. 55 No. 2, pp. 227-246.
Knutson, B.J., Beck, J.A., Kim, S.H. and Cha, J. (2007), “Identifying the dimensions of the experience
254 construct”, Journal of Hospitality Marketing & Management, Vol. 15 No. 3, pp. 31-47.
Ko, E., Kim, E.Y. and Lee, E.K. (2009), “Modeling consumer adoption of mobile shopping for fashion
products in Korea”, Psychology and Marketing, Vol. 26 No. 7, pp. 669-687.
Kwak, N. and Radler, B. (2002), “A comparison between mail and web surveys: response pattern,
respondent profile, and data quality”, Journal of Official Statistics, Vol. 18 No. 2, pp. 257-273.
Ladhari, R., Ladhari, I. and Morales, M. (2011), “Bank service quality: comparing Canadian and Tunisian
customer perceptions”, International Journal of Bank Marketing, Vol. 29 No. 3, pp. 224-246.
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

Lee, K.C. and Chung, N. (2009), “Understanding factors affecting trust in and satisfaction with mobile
banking in Korea: a modified DeLone and McLean’s model perspective”, Interacting with
Computer, Vol. 21 Nos 5/6, pp. 385-392.
Lee, N. and Greenley, G. (2008), “The primacy of theory”, European Journal of Marketing, Vol. 42 No. 9/10,
pp. 873-878.
Levy, S. and Hino, H. (2016), “Emotional brand attachment: a factor in customer-bank relationships”,
International Journal of Bank Marketing, Vol. 34 No. 2, pp. 136-150.
Liang, C., Wang, W. and Farquhar, J.D. (2009), “The influence of customer perceptions on financial
performance in financial services”, International Journal of Bank Marketing, Vol. 27 No. 2,
pp. 129-149.
Maklan, S. and Klaus, P. (2011), “Customer experience: are we measuring the right things?”,
International Journal of Market Research, Vol. 53 No. 6, pp. 771-792.
Martins, C., Oliveira, T. and Popovic, A. (2014), “Understanding the internet banking adoption: a
unified theory of acceptance and use of technology and perceived risk application”, International
Journal of Information Management, Vol. 34 No. 1, pp. 1-13.
Meyer, C. and Schwager, A. (2007), “Understanding customer experience”, Harvard Business Review,
Vol. 85 No. 2, pp. 116-126.
Mols, N.P. (2001), “Organizing for the effective introduction of new distribution channels in retail
banking”, European Journal of Marketing, Vol. 35 No. 5/6, pp. 661-686.
Monferrer-Tirado, D., Estrada-Guillén, M., Fandos-Roig, J.C., Moliner-Tena, M.A. and Sánchez García, J.
(2016), “Service quality in bank during an economic crisis”, International Journal of Bank
Marketing, Vol. 34 No. 2, pp. 235-259.
Novak, T.P., Hoffman, D.L. and Yung, Y. (2000), “Measuring the customer experience in online
environments: a structural modeling approach”, Marketing Science, Vol. 19 No. 1, pp. 22-42.
Nunnally, J.C. (1978), Psychometric Theory, McGraw-Hill, New York, NY.
Parasuraman, A., Zeithaml, V.A. and Berry, L.L. (1988), “SERVQUAL: a multiple-item scale for
measuring consumer perceptions of service quality”, Journal of Retailing, Vol. 64 No. 1, pp. 12-40.
Patsiotis, A.G., Hughes, T. and Webber, D.J. (2012), “Adopters and non-adopters of internet banking: a
segmentation study”, International Journal of Bank Marketing, Vol. 30 No. 1, pp. 20-42.
Peter, J.P. (1981), “Construct validity: a review of basic issues and marketing practices”, Journal of
Marketing Research, Vol. 18 No. 2, pp. 133-145.
Piyathasanan, B., Mathies, C., Wetzels, M., Patterson, P.G. and Ruyter, K. (2015), “A hierarchical model
of virtual experience and its influences on the perceived value and loyalty of customers”,
International Journal of Electronic Commerce, Vol. 19 No. 2, pp. 126-158.
Raza, S.A., Jawaid, S.T., Hassan, A. and Burton, B. (2015), “Internet banking and customer satisfaction
in Pakistan”, Qualitative Research in Financial Markets, Vol. 7 No. 1, pp. 24-36.
Reichheld, F.F. (2003), “The one number you need to grow”, Harvard Business Review, Vol. 81 No. 12, UK customers’
pp. 46-54. perceptions
Reichheld, F.F., Markey, R.G. Jr and Hopton, C. (2000), “The loyalty effect – the relationship between
loyalty and profits”, European Business Journal, Vol. 12 No. 3, pp. 134-139.
Ritter, L.A. and Sue, V.M. (2007), “Introduction to using online surveys”, New Directions for Evaluation,
Vol. 2007 No. 115, pp. 5-14.
Saleem, M.A., Zahra, S., Ahmad, R. and Ismail, H. (2016), “Predictors of customer loyalty in the 255
Pakistani banking industry: a moderated-mediation study”, International Journal of Bank
Marketing, Vol. 34 No. 3, pp. 411-430.
Schmitt, B.H. (2003), Customer Experience Management: A Revolutionary Approach to Connecting with
your Customers, John Wiley & Sons Inc., Hoboken, NJ.
Sharma, S. (1996), Applied Multivariate Techniques, Wiley Publishing, New York, NY.
Sundarraj, R.P. and Wu, J. (2005), “Using information-systems constructs to study online-and
Downloaded by University of Sunderland At 02:01 09 October 2018 (PT)

telephone-banking technologies”, Electronic Commerce Research and Applications, Vol. 4 No. 4,


pp. 427-443.
Tabachnick, B. and Fidell, L. (2001), Using Multivariate Statistics, Allyn and Bacon, London.
Valenzuela, L., Torres, E., Hidalgo, P. and Farías, P. (2014), “Salesperson CLV orientation’s effect on
performance”, Journal of Business Research, Vol. 67 No. 4, pp. 550-557.
Van Looy, B.V., Dierdonck, R.V. and Gemmel, P. (1998), Services Management: An Integrated Approach,
Financial Times, Pitman Publishing, London.
Verhoef, P., Lemon, K., Parasuraman, A., Roggeveen, A., Tsiros, M. and Schlesinger, L. (2009),
“Customer experience creation: determinants, dynamics and management strategies”, Journal of
Retailing, Vol. 85 No. 1, pp. 31-41.
Wu, L.W. (2011), “Beyond satisfaction: the relative importance of locational convenience, interpersonal
relationships, and commitment across service types”, Managing Service Quality, Vol. 21 No. 3,
pp. 240-263.
Yee, R.W.Y., Yeung, A.C.L. and Cheng, T.C.E. (2010), “An empirical study of employee loyalty, service
quality and firm performance in the service industry”, International Journal of Production
Economics, Vol. 124 No. 1, pp. 109-120.
Zhou, T. (2012), “Understanding users’ initial trust in mobile banking: an elaboration likelihood
perspective”, Computers in Human Behaviour, Vol. 28 No. 4, pp. 1518-1525.

Corresponding author
Cajetan I. Mbama can be contacted at: c.mbama@ntlworld.com

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com

Potrebbero piacerti anche